Convergent and sequential synthesis designs: implications for conducting and reporting systematic reviews of qualitative and quantitative evidence